Proteins of the Bcl-2 family are critical for regulating apoptosis and are promising therapeutic targets for treating a wide variety of cancers. Importantly, interaction specificity among pro- and anti-apoptotic family members is known to be critical to how they function. Although high-resolution structures have been solved for several Bcl-2 family complexes, these do not explain why interactions occur between some family members and not others, or how a range of binding affinities is achieved through variations in sequence and structure. A better understanding of the molecular basis for interaction specificity is necessary if we wish to predict, design or disrupt these and other protein interactions using rational methods. The goal of this proposal is to use an integrated program of computational and experimental methods to decipher the interactions of the Bcl-2 family of proteins and to probe their biophysical origins. The specific aims are to: (1) Carry out systematic experimental interaction studies of native, mutant and designed Bcl-2 proteins in order to relate their sequences to their binding properties and provide data for testing and improving computational models, (2) Use computational protein design to identify a wide variety of new peptides that bind to Bcl-2 family members, (3) Develop approaches for combining computational design with experimental selection to engineer peptides with desired interaction specificities (i.e. the ability to interact with some Bcl-2 family proteins but not others) and (4) Use insights into sequence requirements from Aims 1-3 to identify new mammalian and viral Bcl-2 family members that may play a role in apoptosis. Designed or selected peptides from this work could be used as reagents for the dissection of the complex signaling networks that regulate apoptosis, or as leads for the development of more targeted therapies. A longer-term goal is to apply these methods to other proteins with interesting interaction characteristics and an important role in apoptosis. A lasting outcome of this work will be a more comprehensive understanding of the relationship between protein sequence, structure and interaction specificity that will help shape the way we think about molecular recognition in the context of the proteome. PUBLIC HEALTH RELEVANCE Imbalances between pro-life and pro-death Bcl-2 family proteins are important in cancer, and interactions among these proteins provide promising therapeutic targets. The objective of this study is to understand at a high level of detail the molecular basis for the interaction specificity of the Bcl-2 family. This will contribute to our understanding of the biology of regulated cell death, support the development of therapeutic inhibitors and provide insights into the biophysics of protein-protein recognition.